Article pubs.acs.org/jpr
Protein Quantification Using a Cleavable Reporter Peptide Elodie Duriez, Stephane Trevisiol, and Bruno Domon* Luxembourg Clinical Proteomics Center, CRP-Santé, 1 A-B rue Thomas Edison, Strassen 1445, Luxembourg S Supporting Information *
ABSTRACT: Peptide and protein quantification based on isotope dilution and mass spectrometry analysis are widely employed for the measurement of biomarkers and in system biology applications. The accuracy and reliability of such quantitative assays depend on the quality of the stable-isotope labeled standards. Although the quantification using stable-isotope labeled peptides is precise, the accuracy of the results can be severely biased by the purity of the internal standards, their stability and formulation, and the determination of their concentration. Here we describe a rapid and cost-efficient method to recalibrate stable isotope labeled peptides in a single LC−MS analysis. The method is based on the equimolar release of a protein reference peptide (used as surrogate for the protein of interest) and a universal reporter peptide during the trypsinization of a concatenated polypeptide standard. The quality and accuracy of data generated with such concatenated polypeptide standards are highlighted by the quantification of two clinically important proteins in urine samples and compared with results obtained with conventional stable isotope labeled reference peptides. Furthermore, the application of the UCRP standards in complex samples is described. KEYWORDS: mass spectrometry, targeted proteomics, peptide/protein quantification, stable isotope labeled peptides, calibration, standard recalibration, cleavable reporter peptide
■
INTRODUCTION During the past decade, mass-spectrometry-based proteomics has become an essential tool in biological and clinical investigation.1,2 Most protein quantification approaches of biological samples generally involve trypsin digestion of the endogenous proteins followed by a targeted LC−MS-based analysis of signature peptides indicative of the proteins of interest.1 An isotope dilution strategy is frequently employed to achieve precise quantification of the proteins of interest using stable isotope-labeled (SIL) standards.1 Several SIL approaches have been proposed, including: synthetic peptides (SIL peptides), extended-peptides, concatemers of peptides (QconCAT), or full-length proteins (PSAQ).3−7 Recently, a novel type of SIL peptides (differing by the number of neutrons incorporated and using high-resolution accurate mass spectrometry) has been proposed.8 In isotope dilution experiments, the accuracy and reliability of the quantitative assay are directly dependent on the quality criteria of the SIL standards, which include isotope incorporation, chemical purity (isotopic purity), and the actual concentration of the SIL standards.9 SIL peptides (sometimes referred to as AQUA peptides) are frequently used due to their commercial availability and straightforward applicability.10−14 However, it is also known that quantification using SIL peptides can be biased, for example, due to a partial uncontrolled loss of the SIL peptides before addition to the biological samples.15 Several factors can be attributed to a reduction of the initial amounts of SIL peptides. First, the solubilization of the lyophilized SIL peptides may be incomplete.16,17 Second, peptides in solution are prone to nonspecific adsorption to hydrophobic surfaces or can be © XXXX American Chemical Society
subject to a time dependent aggregation. These effects are difficult to predict based solely on the peptide sequence, and thus the formulation buffer, the vials, and storage conditions are critical assay components.18 Thus, the quantification based on SIL peptides often results in precise but inaccurate results, unless the various pitfalls associated with the method are taken into consideration. The LC−MS-based quantification of proteins after digestion relies on the initial amount of the SIL peptides indicated by the manufacturer, which is determined either by photometric methods or by quantitative amino acid analysis (AAA).19 In principle, the SIL peptide concentration should be recalibrated, before each use, to achieve accurate quantitative results. As AAA assays are tedious, time-consuming, costly, and usually performed by an external laboratory requiring significant amount of material (typically >50 μg) for a single analysis, the method is not well-suited for routine recalibration of the SIL peptide concentration. There is a need for methods allowing the recalibration of SIL peptides amount in a routine analysis suitable fashion. Different methods have been recently proposed for the quantification of SIL standards based on equimolar products generation by proteolysis.16,20 Synthetic peptides isotopically labeled at the N-terminus and concatenated to conventional SIL peptides by an enzymatic cleavable site have been proposed to determine the stoichiometry of protein complexes.16 Similarly, the generation of an equimolar mixture upon Received: July 22, 2014
A
dx.doi.org/10.1021/pr500764e | J. Proteome Res. XXXX, XXX, XXX−XXX
Journal of Proteome Research
Article
Figure 1. Representation of the concatenated polypeptide standards and the methodology associated. (A) Individual analyses of the polypeptide standards, incorporating a tryptic cleavage site and containing a protein reference peptide (PRP) ligated to a universal reporter peptide (URP). Because the URP sequence is the same for all UCRP standards, multiplexing PRP calibration is possible using different SIL amino acids of the URP. (B) Analyses of the digested standard samples and determination of the relative response factor between the 1:1 stoichiometric released products, URP and PRP. (C) Determination of the URP and PRP amounts using a reverse URP standard calibration curve.
5%, >97% peptide purity, and >99% isotopic enrichment. Purified peptides were delivered as single-use aliquots in glass tubes.
digestion of tandem peptides was used to determine relative MS ionization efficiencies.20 Full-length protein standards containing an N-terminal enzymatic cleavable peptide have been proposed where the N-terminal peptide, as a surrogate for the whole protein, is quantified.21 The use of concatenated synthetic polypeptides with a trypsin cleavage site can be further exploited to systematically determine the concentration of the SIL peptide standard of interest upon tryptic digestion. (See Figure 1.) The universal cleavable reporter peptide (UCRP) standard discussed here has the peptide to be quantified fused to a universal reporter peptide (URP), whose concentration can be readily determined and used to calculate the relative response of the protein reference peptide (PRP). Conversely, if the response factors have already been determined, the concentration of the URP allows us to determine the concentration of the PRP of interest. As the URP sequence is identical for all UCRP standards, this peptide can be systematically used to recalibrate any PRP (Figure 1). The methodology described in this study includes the determination of the PRP concentration, using a calibration curve composed of URP isotopologues accurately quantified, and the comparison of the cleavable reporter peptide standards to the results obtained using direct quantification with conventional SIL peptides. The calibration using UCRP standards can be performed either individually or concomitantly if multiple labeling schemes are used.
■
Trypsinolysis of the UCRP Standards
Each UCRP standard (1.5 nmol) was reduced with 6 mM dithiothreitol (Sigma-Aldrich, St. Louis, MO) in 50 mM ammonium bicarbonate (Sigma-Aldrich) at 50 °C for 45 min and then alkylated with 42 mM iodoacetamide (Sigma-Aldrich) in 50 mM ammonium bicarbonate at room temperature for 45 min. The reduction and alkylation steps can be omitted for the digestion of the pure UCRPs because they do not contain free cysteine residues. The pH was adjusted to 8.5 with 1 M NaOH (SigmaAldrich) and the sample was digested overnight with trypsin (sequencing-grade trypsin, Promega, Madison, WI) using a ratio of 1:20 (w/w) at 37 °C. The digested sample was desalted using Sep-Pak C18 reverse-phase cartridges (Waters, Milford, MA). Peptides were eluted using 1 mL of 70% acetonitrile (CHROMASOLV Plus, Sigma-Aldrich) in 0.1% formic acid (Sigma-Aldrich) and dried using a vacuum concentrator. The dried samples were stored at −20 °C until LC−MS analysis. External and Direct Calibration in a Single LC−MS run
Ten synthetic SIL peptide variants of the AALPAAFK URP sequence with various combinations of 15N- and 13C-labeled amino acids (labeling underlined) were synthesized and solutions of different concentrations were prepared (2 amol/ μL AALPAAFK, 4 amol/μL AALPAAFK, 14 amol/μL AALPAAFK, 41 amol/μL AALPAAFK, 124 amol/μL AALPAAFK, 370 amol/μL AALPAAFK, 1.1 fmol/μL AALPAAFK, 3.3 fmol/μL AALPAAFK, 10.0 fmol/μL AALPAAFK, and 30.0 fmol/μL AALPAAFK). These solutions were spiked: (i) with a UCRP standard previously digested with trypsin for external calibration and (ii) into a biological sample containing an
MATERIAL AND METHODS
Synthetic Peptides
Isotopically labeled (SIL) peptides, URP isotopologues, and UCRP standards were synthesized in crude form by Thermo Fisher Scientific (Ulm, Germany). HPLC purification yielded peptides with a concentration precision equal to or better than B
dx.doi.org/10.1021/pr500764e | J. Proteome Res. XXXX, XXX, XXX−XXX
Journal of Proteome Research
Article
HB), and then desalted on C18 cartridges (elution with 0.1% formic acid/50% acetonitrile/water) and dried using a vacuum concentrator. The sample was solubilized in 0.1% formic acid to obtain a final concentration of 1 μg/mL. LC−SRM analysis was performed on a triple quadrupole instrument as described later.
undigested UCRP standard. LC−SRM analysis was performed on a triple quadrupole instrument as described later. Preparation of Biological Samples
Collection of Urine Samples. Pooled human urine was provided by the Integrated Biobank of Luxembourg (IBBL). Midstream urine samples were collected from 10 nonsmoking healthy volunteers, five females and five males, aged between 30 and 40 years. There was no history of renal dysfunction in any of the subjects, and the individuals were not medicated at the time of sample collection. Urine samples were centrifuged at 1000g for 20 min at room temperature. The supernatants were pooled and stored as 50 mL aliquots in falcon tubes at −80 °C. Treatment of Urine Samples. The amount of urinary proteins was determined by the pyrogallol assay (SigmaAldrich). Samples corresponding to ∼250 μg of urinary protein were precipitated overnight with acetonitrile at a ratio of 1:5 (v/v). After centrifugation at 14 000g for 30 min at 4 °C, the pellets were washed once with acetonitrile, air-dried, and suspended in 250 μL of 8 M urea (Sigma-Aldrich) in 0.1 M ammonium bicarbonate. The samples were reduced with 20 mM dithiothreitol in 50 mM ammonium bicarbonate at 37 °C for 30 min and alkylated with 80 mM iodoacetamide in 50 mM ammonium bicarbonate at 37 °C for 30 min. Sample volumes were adjusted to reach a 2 M urea concentration using 100 mM ammonium bicarbonate. Samples were then digested overnight at 37 °C with trypsin at a 1:20 (w/w) ratio, and digestion was stopped by the addition of formic acid to reach a pH 2. Sep-Pak C18 reverse-phase cartridges were used to clean up and desalt the samples after digestion; the peptides were eluted using 1 mL of 50% acetonitrile in 0.1% formic acid and dried using a vacuum concentrator. The dried samples were stored at −20 °C until the LC−MS analysis. For the preparation of dilution series, dried urine samples were solubilized in 0.1% formic acid at a final concentration of 1 μg/μL. Five SIL peptides (DGAGDVAFVK, SASDLTWDNLK, EGYYGYTGAFR, LLLTSAPSLATSPAFR, YDLLDLTR) were independently spiked into the urine digest to prepare in parallel two types of samples: first, an urine digest containing calibrated SIL peptides derived from the UCRP standards (0, 0.02, 0.05, 0.1, 0.4, 0.9, 2.3, 5.9, 15.2, and 39.4 fmol/μL) before applying a correction after recalibration based on the response factor, and second, a urine digest containing freshly prepared, conventional, SIL peptides (0, 0.02, 0.05, 0.1, 0.4, 0.9, 2.3, 5.9, 15.2, and 39.4 fmol/μL), with both C-terminal 15N and 13Clabeled arginine and lysine residues (labeling underlined). Collection and Treatment of Plasma Samples. Blood serum obtained from patients diagnosed with lung cancer was provided by the Integrated Biobank of Luxembourg (IBBL). The two most abundant proteins (human albumin and IgG) were depleted using a multiple affinity removal spin cartridge (MARS 2, Agilent Technologies). Protein concentrations were measured by the Bradford assay (Sigma-Aldrich) before and after the depletion process. After depletion, the sample was supplemented with the first analogue of a UCRP standard acting as a surrogate of human SAA-1 protein, GPGGVWAAEAISDARAALPAAKF (S1−HA), and then reduced with 10 mM dithiothreitol at 50 °C for 50 min, alkylated with 25 mM iodoacetamide in the dark at RT for 30 min, and digested with trypsin using a ratio of 1:20 (w/w). The sample was supplemented with the URP calibration mixture and the calibrated PRP from the second analogue of the UCRP standard of interest, GPGGVWAAEAISDARAALPAAFK (S1−
Liquid Chromatography and Mass Spectrometry
LC Separation. All peptide separations were carried out on a Dionex Ultimate 3000 RSLC-nano system (Thermo Scientific). For each analysis, the sample was loaded into a Dionex Acclaim PepMap trap column (2 cm × 75 μm i.d., C18, 3 μm, 100 Å) at 5 μL/min using an aqueous solution of 0.05% (v/v) trifluoroacetic acid (Sigma-Aldrich) and 1% acetonitrile. After 3 min, the trap column was set online with a Dionex Acclaim PepMap RSLC analytical column (15 cm × 75 μm i.d., C18, 2 μm, 100 Å). Peptide separation was performed by applying a mixture of solvent A/B. Solvent A was HPLC-grade water with 0.1% (v/v) formic acid, and solvent B was HPLCgrade acetonitrile with 0.1% (v/v) formic acid. Separations were performed by applying (i) a linear gradient of 2−35% solvent B in solvent A at 300 nL/min over 48 min, followed by a washing step (5 min at 90% solvent B in solvent A) and an equilibration step (10 min at 2% solvent B in solvent A) or (ii) a stepwise gradient of 17% solvent B in solvent A over 5 min, followed by a washing step (4 min at 90% solvent B in solvent A) and an equilibration step (10 min at 2% solvent B in solvent A). Sample injection volume was 1 μL. Analyses on a Quadrupole-Orbitrap Instrument. SIM and PRM analyses were performed using a Q-Exactive mass spectrometer (Thermo Scientific, Bremen, Germany). A dynamic nanoelectrospray source was utilized with uncoated silica tips of 12 cm length, 360 μm outer diameter, 20 μm inner diameter, and 10 μm tip inner diameter. For ionization, 1500 V of liquid junction voltage and capillary temperature of 250 °C were used. For the analyses of the dilution series of five peptides (calibrated SIL peptides derived from the UCRP standards or conventional SIL peptides) in urine samples (performed in triplicate), the acquisition method combined two scan events corresponding to a full-scan method and a timescheduled sequential PRM method targeting the five pairs of SIL peptides/endogenous peptides in ±1 min retention time windows. The full-scan method employed a m/z 300−1500 mass selection, an Orbitrap resolution of 70 000 (at m/z 200), a target automatic gain control (AGC) value of 1 × 106, and maximum fill times of 250 ms. The time-scheduled PRM method employed an Orbitrap resolution of 35 000 (at m/z 200), a target AGC value of 1 × 106, and maximum fill times of 120 ms. The precursor ion of each targeted peptide was isolated using a 2 m/z unit window. Fragmentation was performed with a normalized collision energy of 25 eV, and MS/MS scans were acquired with a starting mass of m/z 100, with the ending mass being automatically defined by the m/z and the charge state of the precursor ion. Data analysis was performed using Pinpoint (version 1.2 Thermo Fisher Scientific). Ion chromatograms were extracted with a mass tolerance of 10 ppm for SIM data and 20 ppm for PRM data. Analyses on a Triple−Quadrupole Instrument. Selected reaction monitoring analyses were performed using a TSQ Vantage extended mass range triple−quadrupole mass spectrometer (Thermo Scientific, San Jose, CA) with identical nanoelectrospray and chromatographic settings as previously described. The selectivity for both Q1 and Q3 quadrupoles was set to 0.7 Da (fwhm). The argon collision gas pressure in the C
dx.doi.org/10.1021/pr500764e | J. Proteome Res. XXXX, XXX, XXX−XXX
Journal of Proteome Research
Article
MS evidence, including peptide elution profiles, ionization efficiency, and fragmentation pattern. Then, preliminary experiments to assess tryptic proteolysis were performed on a series of UCRP standards created by fusing 16 distinct URP sequences (SI, Table S-1) to a single PRP. Three PRPs were carefully selected based on their amino acid composition as positive and negative controls for proteolysis (taking into account the fusion at the N-or-C-termini) based on the miscleavage peptide patterns established by Thiede et al.27 The sequences of the three PRPs were SFFSFLGEAFDGAR, ELDESLQVAER, and ASSIIDELFQDR, respectively labeled PRP1, PRP2, and PRP3. The cleavable reporter peptide standards were digested by trypsin at 37 °C for 12 h. A desalting step was performed before injection in the LC−MS system. The efficiency of the enzymatic digestion was found to be optimal by positioning the PRP in front of the URP. This is consistent with previous reports27,28 indicating that trypsin cleaves less efficiently after lysine residues as compared with arginine residues. Out of those UCRP standards with PRPs in front of the URP, three URP sequences ([AALPAAFK], [AANFAAFK], and [AAQLAALK]) showed the best efficiency of proteolysis for the three combinations of resulting concatenated polypeptides (PRP1-URP, PRP2-URP, and PRP3-URP). More specifically, the proteolysis by trypsin was nearly complete (>99%) for the UCRP standards containing the AALPAAFK URP sequence. From this set of experiments, the position of the URP was fixed C-terminal in the resulting cleavable standards, and the AALPAAFK URP sequence was retained for further investigation. To better assess the proteolysis efficiency, we performed a second set of experiments on a new set of UCRP standards containing the selected URP sequence: AALPAAFK. It was fused to 43 additional PRPs (PRP4-URP to PRP46-URP) for UCRP standard synthesis (SI, Table S-2). The efficiency of proteolysis obtained with the AALPAAFK URP sequence for the 43 PRPs was determined to be 95.8% (CV 8.9%), demonstrating that the selected URP can be cleaved off the UCRP standards regardless of the PRP sequence. Finally, the trypsinization efficiency of the UCRP standard was determined using a SIL analogue (identical sequence with different SIL amino acids) of the standard. An example is illustrated in Figure 2. Undigested and trypsin digested samples (Samples 1 and 2, respectively) of the UCRP standard GPGGVWAAEAISDARAALPAAFK were spiked with a defined amount of the undigested SIL analogue GPGGVWAAEAISDARAALPAAFK (labeling underlined). LC−SRM analysis monitoring the UCRP standard, its isotopologue, and the trypsin digestion products (both PRP and URP) was performed in triplicate (Figure 2). The UCRP standard was found to be very effectively digested by trypsin. The SRM measurements were sufficiently accurate to determine a trypsin digestion rate of >99% as the S/N ratio of the residual undigested standard was lower than 1% of the S/N ratio of the standard prior digestion (Figure 2). Following this methodology, the efficiency of the trypsin digestion of the UCRP standard can be precisely determined. However, it is important to notice that due to the resulting inherent equimolar amount of both peptides upon enzyme digestion, (i.e., PRP and URP), an unexpected incomplete proteolysis of the UCRP standard is not of relevance for subsequent calibration.
second quadrupole Q2 was set at 1.5 mTorr. For each peptide, the selection of the monitored transitions and the optimization of the collision energy required were performed as described previously.22 Data analysis was performed using Pinpoint (version 1.2 Thermo Fisher Scientific) or Skyline. Quantification Based on Isotope Dilution Strategy
For the analyses of dilution series using SRM or PRM methods, the area under the curve (AUC) of each targeted transition (SRM analysis) and selected fragment ion (PRM analysis) was determined for each dilution point. For the analyses of the dilution series of URP isotopologues in a single LC−SRM run, the peptide AUCs were directly used to establish the corresponding dilution curves. For the analyses of the dilution series of the five SIL peptides in urine samples by PRM analysis, the peptide AUCs were employed to calculate SIL/ endogenous peptide AUC ratios. These SIL/endogenous peptide AUC ratios were then used to establish the dilution curves of each peptide. For each dilution series, a linear regression analysis was performed. The range of linearity was defined as the range of spiked peptide amounts for which the relative difference between calculated concentrations and the spiked concentrations was
